Houdini 2.0c Pro x64 - Stockfish 2.2.2 JA SSE42 Match Results

1   Houdini 2.0c Pro x64 1CPU          +35/=51/-14 60.50%   60.5/100
2   Stockfish 2.2.2 JA SSE42 1CPU  +14/=51/-35 39.50%   39.5/100

Download the games here: http://goo.gl/S1CcM

Komodo64 4 SSE - Rybka 4.1 SSE42 x64 Match Results

1   Komodo64 SSE Version 4 1CPU   +29/=52/-19 55.00%   55.0/100
2   Rybka 4.1 SSE42 x64 1CPU          +19/=52/-29 45.00%   45.0/100

What was interesting to me is that Komodo was +9 over Rybka in the first 50 games and Houdini was drawn vs Rybka at that point. Even so, Houdini finished with the better score (+12 vs +10).

Download the games here: http://goo.gl/2zuoK

Critter 1.4a x64 SSE4 - Komodo64 4 SSE Match Results

1   Komodo64 SSE Version 4 1CPU    +28/=49/-23 52.50%   52.5/100
2   Critter 1.4a 64-bit SSE4 1CPU        +23/=49/-28 47.50%   47.5/100

This match went back and forth, and it was fun to watch.

Download the games here: http://goo.gl/ap4rq

Houdini 2.0c Pro x64 - Rybka 4.1 x64 SSE42 Match Results

1   Houdini 2.0c Pro x64 1CPU     +36/=40/-24 56.00%   56.0/100
2   Rybka 4.1 SSE42 x64 1CPU   +24/=40/-36 44.00%   44.0/100

If the match were to have stopped half way through, it would have been in a draw. Houdini got its +12 score in the final 50 rounds.

Download the games here: http://goo.gl/7I2f3

Houdini 2.0c Pro x64 - Komodo64 4 SSE Match Results

1   Houdini 2.0c Pro x64 1CPU           +32/=48/-20 56.00%   56.0/100
2   Komodo64 SSE Version 4 1CPU  +20/=48/-32 44.00%   44.0/100

Download the games here: http://goo.gl/ZdC86

Critter 1.4a SSE4 - Stockfish 2.2.2 JA Match Results

1   Critter 1.4 64-bit SSE4 1CPU        +25/=53/-22 51.50%   51.5/100
2   Stockfish 2.2.2 JA SSE42 1CPU   +22/=53/-25 48.50%   48.5/100

Stockfish won the last two rounds (99 & 100) as White and Black in the Kings Indian, Classical, to end with a -3 score. In game 99, around move 80, Stockfish had Black and was running only on increment time (compared to over 2:00 left for Critter, if I remember correctly), yet it was able to turn the game and pull out a win. It was fun to watch.

Download the games here: http://goo.gl/AB9Dl

Komodo64 4 SSE vs Stockfish 2.2.2 JA SSE42 Results

1   Komodo64 SSE Version 4 1CPU  +29/=54/-16 56.5%   56.5/100
2   Stockfish 2.2.2 JA SSE42 1CPU   +16/=54/-29 43.5%   43.5/100

Download the games here: http://goo.gl/sR4SU

Critter 1.4a SSE4 - Houdini 2.0c match results

1   Houdini 2.0c Pro x64 1CPU        +30/=47/-23 53.50%   53.5/100
2   Critter 1.4a 64-bit SSE4 1CPU   +23/=47/-30 46.50%   46.5/100

Download the games here: http://goo.gl/uds9p

Testing Control Change

I am changing my time control to 10' + 10" increment (10 minutes per game + 10 seconds per move) and starting a new ratings database. I am also giving each engine 64 MBs of endgame cache, instead of 32. I think this is a respectable time control and is long enough to produce quality games. The standard benchmark for other chess engine test groups is based on an AMD X2 4600+ @ 2.4 GHz. My machine benchmarks almost twice as fast as that, using Crafty Bench 19.17. My computer processes the same amount of data in 40/21 as a "standard" PC does in 40/40. The total thinking time each engine of mine will have in a typical 80-move game is 2,800 seconds, or about 47 minutes. If I were to participate in a 40/40 list, my computer's thinking time would be half of that, which is 21 minutes for 40 moves (or 42 minutes for 80 moves). Forty-two minutes = 2,520 seconds, which is still  about 300 seconds less than the 2,800 seconds my engines will think in 80 moves. If any game goes beyond 80 moves, the quality should not suffer much: There is still the left-over 300 seconds plus 10 seconds per move, which, usually in an endgame, is plenty of time as to not affect game quality. Also, ponder is on, so it could be as much as 20 seconds/move thinking time.

Critter 1.4a x64 SSE4 - Houdini 2.0c x64 Match Results

ATTENTION: This match was ran at the old time control of 5' + 5''. This entire match was ran again with the new time control, and none of these games are included in the ratings report.

1   Houdini 2.0c Pro x64 1CPU        +34/=52/-14 60.00%   60.0/100
2   Critter 1.4a 64-bit SSE4 1CPU   +14/=52/-34 40.00%   40.0/100

I am a bit surprised by this result: Houdini wins by 20. Critter beat Komodo by 1, and Houdini only beat Komodo by 3, so I am surprised, a bit, by the margin in this match. I mean no offense to either engine; I just expected it to be closer. Nothing seems amiss in the PGN file. All the controls were the same. The PC was unplugged from the Internet, and all extra processes were stopped, just like all the other games. I watched many games, and nothing seemed wrong; CPU utilization was the same, no strange processes were running, the anti-virus scan never launched, etc. I don't think it is because I am using Critter 1.4a (which I have used throughout these games that I am including in the rating and is meant to fix issues with engine matches with ponder = on), because IPON uses the same version and it's quite strong in their ratings.

Download the games here: http://goo.gl/naozd

Critter 1.4a x64 SSE4 - Houdini 2.0c x64

The Critter 1.4a x64 SSE4 vs. Houdini 2.0c x64 match just started. Updates will be on Twitter.

Critter 1.4a x64 SSE4 - Komodo64 4 SSE Results

ATTENTION: This match was ran at the old time control of 5' + 5''. This entire match was ran again with the new time control, and none of these games are included in the ratings report.

1   Critter 1.4a 64-bit SSE4 1CPU          +24/=53/-23 50.50%   50.5/100
2   Komodo64 SSE Version 4 1CPU      +23/=53/-24 49.50%   49.5/100

Critter was up by 3 early on, then it was down by 4, then it was up by 3, and finally the score was almost a dead-draw. Critter won by 1 game.

Download the games here: http://goo.gl/9JDgp

Critter 1.4a x64 SSE4 - Komodo64 4 SSE

The match of Critter 1.4 x64 SSE4 vs Stockfish 2.2.2 x64 JA SSE42 was stopped prematurely. It has come to my attention that Richard Vida, the maker of Critter, released a new version of Critter to fix issues for people running engine matches with ponder = on. See this.

Also, for rating purposes, I need to have the engines be able to measure themselves against something. So, Komodo will be the connector in this game, since it played Houdini in the last match. Critter 1.4a x64 SSE4 vs Komodo64 4 SSE just started. Updates as the match progresses will be on Twitter.

Happy chessing!

Critter 1.4 x64 SSE 4 - Stockfish 2.2.2 JA SSE42

The next match will begin this afternoon (Chicago time), and it will feature Critter 1.4 x64 SSE4 vs Stockfish 2.2.2 JA SSE42.

Houdini 2.0c x64 - Komodo64 4 SSE results

ATTENTION: This match was ran at the old time control of 5' + 5''. This entire match was ran again with the new time control, and none of these games are included in the ratings report.

 
1   Houdini 2.0c Pro x64 1CPU            +28/=47/-25 51.50%   51.5/100

2   Komodo64 SSE Version 4 1CPU   +25/=47/-28 48.50%   48.5/100

Very interesting match with 53 decisive games.

Download the games here: http://goo.gl/uymnn

ATOMICC Testing on Twitter

I am now on Twitter. I will be posting status updates as the games progress on Twitter, but not on my blog. Follow the games @ATOMICCTesting

Houdini 2.0c x64 vs Komodo64 4 SSE

A 100-game match between Houdini 2.0c x64 vs Komodo64 4 SSE is in progress. Results will be posted soon.

Welcome to ATOMICC testing

All The Options Matter In Computer Chess.

I do my best to create a level playing field in my engine tests, and take it quite seriously. Thanks for your interest.

All time-stamps are London time (GMT).

:)